Project Summary/Abstract In the United States, the joint disease osteoarthritis (OA) debilitates over one-third of people over sixty-five years old and causes hundreds of thousands of knee replacements annually. Despite its high prevalence, little is known about the molecular mechanisms that regulate knee formation and OA risk and how one inherits risk at specific joints (e.g., knee versus hip). Recent GWAS have identified at least seventeen loci that significantly associate with knee OA risk. However, the casual variants for these loci have not been identified because their association signals span large genomic intervals harboring uninvestigated non-coding regulatory regions. Of these, common variants in the Growth Differentiation Factor Five gene (GDF5), a critical regulator of joint development, reproducibly associate with knee OA risk in human populations. GDF5 OA variants reside on a high frequency 130 kb haplotype possessing numerous mutations that each may be causal for OA risk, but interestingly, no protein coding mutations have been uncovered that explain the associations. In a complementary study, we discovered ten GDF5 regulatory elements (e.g., promoters, enhancers) spanning this interval and revealed that they function with tremendous joint specificity (e.g., knee vs. hip). These enhancers were initially tested at incipient stages of GDF5 expression in mouse embryos, but we also know that GDF5 contributes to the differentiation of knee structures (ligaments, tendons, menisci, articular surfaces) well after this stage of development. Our research will bring together knee OA risk variants in patient populations and our knowledge of how GDF5 is controlled at the DNA level. We first aim to assess the functional contributions of joint-specific GDF5 enhancers to pre-natal knee development and adult joint homeostasis using CRISPR-Cas9 editing to excise these elements in vitro in human cells and in vivo in the mouse. Strikingly, our pilot analyses also revealed that a subset of these GDF5 enhancers possesses common and rare human variants in strong linkage disequilibrium with the highest associated OA variants in the interval. Our second aim is to further explore the association between genetic variants in the locus and OA knee shape using data acquired from the complete Osteoarthritis Initiative MR database on adult knees. Our preliminary data reveal that several variants associated with OA knee shape reside in several functional enhancers that control expression of GDF5 in the knee. Finally, our third aim is to functionally test these human regulatory variants for their impact on enhancer activity, knee formation, and OA risk by using transfection studies and CRISPR-Cas9 in human cartilage cells and in the mouse model. Completion of these studies will reveal functional variants in GDF5 that underlie its role in knee shape and OA risk.